Moderate length peptides have attracted considerable research and commercial interest by virtue of the properties some exhibit in enhancing, blocking or otherwise affecting the activity of receptors, microbes, and other molecules deemed biologically significant. Specifically, hexapeptides have proven to have a sufficient chain length to block much larger molecules such as receptors, enzymes and antibodies. Thus, synthetic and natural hexapeptides have exhibited diverse therapeutic properties, among them: Antimicrobials with minimum inhibitory concentrations an order of magnitude less than known natural antimicrobial peptides; bactericides; antivirals; activity as antigenic determinants; and the like. The problem is that there are 64 million (64 m) hexapeptide combinations for the twenty L-amino acids, and another 64 m for the D-amino acids. Indeed if the selection were made from all of the L and D combinations the number amounts to 4.096 billion. Since there are in turn millions of biologically/medically significant targets, preparing a complete suite of just 64 m L-hexapeptides and assaying activity for each of the millions of targets is, practically speaking, an infinite, and therefore, impossible, task.
Accordingly, the Synthetic Peptide Combinational Library (SPCL) approach has recently resulted in a manageable approach to the problem of screening for a unique hexapeptide among the 64 m that is the most active for a given target. In order to be feasible, libraries of large numbers of hexapeptides, on the order of 100,000 or so at a time, must be prepared in quantities sufficient to result in a positively determinable reaction.
There are currently four basic library techniques offered: viral approachs (originated by George Smith of LSU, and by Cetus and Affymax independently); the Houghten approach using Tea-bags; the Selectide bead approach; and the Affymax Chip approach. The latter three have distinct advantages over the viral approach in which peptide libraries are displayed by bacteriophages (viruses that prey on bacteria). A short degenerate oligonucleotide encoding all combinations of a short peptide sequence is cloned into Gene III or XIII of a filamentous phage and expressed on the phage surface. Recombinant phage are screened with the target molecule (e.g. receptor), and phage expressing a certain peptide that binds to the target are identified. Nucleotide sequence analysis of the recombinant Gene III or Gene XIII identifies the peptide sequence displayed by the binding phage.
The problem with the viral approach is that the range of peptides is limited to those tolerable by virus and E. Coli. That is, only a limited suite of peptides can be produced from among the 64 m possible hexapeptides, and likewise for the even greater numbers of longer peptides. Additionally, only L- amino acids are allowed, and each individual hexapeptide of the library is produced within the phage as fusion products. This reduces the flexibility of the sequences, and may mask them entirely.
The Selectide bead approach uses vast quantities of spherical crosslinked polymer beads (Millipore or Cambridge Research Laboratories polyacrylamide beads or Rapp Tentagel polystyrene) divided into 20 equal piles, each of which then has a different L-amino acid coupled to all the beads in the pile. The bead piles are then combined and thoroughly mixed. The resulting single pile is again divided into 20 different piles, each of which is reacted with a different one of the 20 different L-amino acids. This Divide, Couple and Recombine process (DCR) is repeated through six reactions to produce hexapeptides bound to the beads. The beads are then screened against a "target" molecule which is marked with a conjugated enzyme, such as horseradish peroxidase. The target "sticks" to active hexapeptide(s). The bead is rendered visible by adding a substrate for the enzyme which converts it to a colored dye which is precipitated within the beads, and then the visually identified bead(s) are picked out with tweezers. The peptides on the beads are then analyzed, for example by the Edman sequencing method, and soluble versions produced in a synthesizer. The initial screening (locating the target bead(s)) takes only days, the makeup of each identified hexapeptide is unknown, and the analysis and synthesis for confirmation and further work takes much longer.
The Houghten (Iterex) Tea-Bag method, shown in U.S. Pat. No. 4,631,211, employs methylbenzhydrylamine (MBHA) polystyrene beads in a number of foraminous containers, e.g. porous polypropylene bags (Tea-Bags), to prepare a truncated SPCL. In order to shorten the processing time, the Tea-Bag process employs partially known, partially undetermined hexapeptide sequences in repeated screenings, followed by iterative resynthesis to replace the unknown AA sequence positions with known AAs, i.e., A-O.sub.1 O.sub.2 O.sub.3 XXX, A-O.sub.1 O.sub.2 O.sub.3 O.sub.4 XX, etc. The method works on the assumption that a biologically significant response can be detected from a solution which contains hundreds of thousands of inactive components.
The Tea-Bag process typically uses 18 of the 20 L-AAs (cysteine and tryptophane are omitted in the initial library for ease of synthesis), starting with 104,976 combinations of non-determined tetrapeptide resins (XXXX-peptide resins) in 324 aliquots, and adds the 324 known dipeptide sequences (18.sup.2) in the terminal two positions. For epitope determination of antibody binding, the 324 pools are screened to see which best inhibits binding of the target antibody with its natural antigen. The most active amino terminal dipeptide sequences are then incorporated into a further set of 20 pools in which the third residue is varied. These are rescreened for low IC. The most active sequences are again reincorporated iteratively to define positions 4-6 to finally obtain a characterized active hexapeptide.
The Tea-Bags employ MBHA-styrene beads and standard t-Boc chemistry (the conventional Merrifield method) in combination with simultaneous multiple peptide synthesis (SMPS) to prepare the starting 18.sup.4 non-determined XXXX-tetrapeptide library by a DCR process, which assures equimolarity of the peptides on the resin. Briefly, 18 porous polypropylene packets, each containing 4.65 mmol (5.00 g) of MBHA resin, are coupled with each of the protected N-x-t-Boc amino acids. Coupling reactions are checked to ensure they are complete (&gt;99.5%) as assessed by Gisin's picric acid or Kaiser's tests. The resulting resins are then combined and thoroughly mixed as with the Selectide bead process. The resulting resin mixture is separated into 18 portions of equal weight which are placed into porous polypropylene packets, followed by N-a-t-Boc protecting group removal and neutralization of the resulting amine TFA salts. The resin packets are then reacted with solutions of the individual activated amino acids to yield the 324 dipeptide combinations (18.sup.2). The above DCR process is repeated twice more, yielding a final mixture of 104,976 protected tetra-peptide resins (18.sup.4). This XXXX-resin is divided into 324 aliquots (150 mg each) and placed in numbered porous polypropylene packets. Synthesis of the next two defined positions is carried out by SMPS. The peptide mixtures are deprotected and cleaved from their respective resins using low-high hydrogen fluoride (HF) in a multiple HF cleavage apparatus (Multiple Peptide Systems, San Diego, Calif.). Extraction of the individual peptide mixtures was carried out with H.sub.2 O. The competitive ELISA used is a modification of the direct ELISA technique, differing only in the antibody additions step in which 25 microliters each peptide mixture of the SPCL was added with a fixed dilution of the antibody (25 microliters per well).
The foraminous container of the Tea-Bag must retain the solid phase beads, yet have a sufficient number of openings to permit ready entrance and exit of solvent and solute molecules at the reaction temperature, but bar exit of the solid phase. While the synthesis is the standard Merrifield technique, new linking groups that attach the X.sub.n -peptide to the styrene bead supports are disclosed. This process can be characterized as not calling for a continuous support, and it is not addressable.
The Affymax "chip" approach is described in PCT publication W090/10570, which discloses a method for multiple peptide synthesis on a solid support. Photolabilely-blocked amino groups are chemically attached (bonded) to a silicon chip, then irradiated through a patterned mask to selectively remove the blocking groups in a pre-arranged pattern. An amino acid will bond by addition only to the irradiation exposed areas. Additional masks are imposed and radiation applied as a prelude to adding second amino acids. Each amino acid added can include a blocking group so that further addition to that site occurs only after irradiation unblocking. Repeating the process with plural masks builds location specific polypeptides. When the chip is exposed to the target molecule, it may stick to one or more locations. By checking coordinates on a map of the chip, the peptide is identified. However, this process does not work with target molecules stuck to, or part of, cells, and there are exposure problems during processing, i.e., some AA's are light sensitive and cannot be used. Further, the reactions at the surface are not complete; for example, where reaction completion is only 90%, by the 6th iteration to obtain a hexapeptide, only half of them will be made properly.
Accordingly, there is a need in the art for a peptide synthesis and screening process that is rapid and accurately identifies the active peptides from amongst those in an extended, reusable SPCL.